PassPredictR: Contextualizing NFL Throwing Decisions Through Modeling Receiver Choice

Zachary Pipping

University of Florida

Lou Zhou

Rice University

Dr. Karim Kassam

Teamworks, Outside Advisor

Motivating Example - PJ Walker to DJ Moore

Motivating Questions

  • What would other quarterbacks do in this situation?
  • Which quarterbacks deviate from the expected option the most?
    • Do some quarterbacks find success in this deviation?
    • Should some quarterbacks throw more typical passes?
  • Look to build a ranking model to determine the most likely throw target

Data Overview

  • 2024 NFL Big Data Bowl – 2022 Season Weeks 1–9
    This text will automatically adjust its size to fit within the 50% width of this column. ::: fragment
  • Game and Play Data – Teams, Score, Play Description, Game Context, Play Result, Changes in Win Probability
  • Player Play Data – Statistics for each player for a play
    • Route ran by player, Whether the player made a tackle or interception
  • Tracking Data - Locations of players and the football at each frame of a play
  • Exclusively looking at throwing plays with an obvious target
    • Removing spikes and throwaways

:::::::::

Current Spacing Tells an Incomplete Story

Speed and Orientation as a Proxy for Future Separation


Deriving QB Line of Sight Prior to Throw

Methodology

  • Building a ranking algorithm(i.e. XGBoost) using hand-crafted features to rank the likeliest recipient at a frame - 59.9 \(\pm\) 0.5% top-1 accuracy
    • Using a random hyperparameter search and 5-fold cross validation, with folds on matches
    • Performs significantly stronger than naive random guess(20%) and choosing the player who is farthest from their closest defender(31%)
  • Applying model to contextualize individual QB decisions by comparing them to model-predicted choices

Feature Set

Feature Category Features
Recipient Features - Distance (x, y, magnitude)*
- Speed Differences (x, y, magnitude)*
- Orientation Differences*
- Speed Vector Distance*
- Receiver Position
- First Down Indicator
- Number of Defenders in Route-Runner’s “Next 5 Yards”
- Angle between QB Orientation and Receiver 5 Frames Prior
Quarterback Features - Distance from Receiver
- Movement Vector
- Under Pressure Indicator
Game Context - Quarter
- Down and Distance
- Score Differential
- Time Remaining



*Feature taken relative to the top-3 closest defenders

For P.J. Walker, The Expected Play is the Safe Play



Checking with the Eye Test



Potential Optimization Opportunities with More Conventional Passes

Discussion

  • Able to model the likely target using an XGBoost Ranking Model with strong predictive power
    • Improvements from incorporating newer factors like receiver skill and pre-snap factors
  • Reliance on proxies which could add bias to results
    • Smaller dataset, only about 200 throws per quarterback
    • Would like to model quarterback development throughout their career

Further Information

Appendix A - Projecting Future Locations With a Point Estimate

::: {style=“text-align: center;} :::

Appendix B - Variable Importance

Appendix C - Example for Vision Features

Appendix D - Optimal QB Vision Derivation

Appendix E - Head Tracking Proof of Concept